- True, but my point is that not only does the analogy not work, the author also doesn't understand the thing he makes the analogy with, or at least explores the thought so shoddily that it makes no sense.
It's somewhat like saying cars are faster than motorbikes because they have more wheels-- it's like with horses and humans, horses have four legs and because of that are faster than humans with two legs. It's wrong on both sides of the analogy.
- [put my hand up]. I recently had to/wanted to convert my lecture slides from latex-beamer/lyx into html/reveal.js. I did a couple of slides per hand, and then asked AI to convert the rest, following my example. Saved me hours of tedious and boring work.
- Fewer absolute or relative? If you scale down your search space... This only makes some kind of sense if your step size is fixed. While I agree with another poster that a reduction of a creative process to gradient descent is not wise, the article also misses the point what makes such a gradient descent hard -- it's not sharp peaks, it's the flat area around them -- and the presence of local minima.
- I think it's a trick. It seems to be the article is just a series of ad-hoc assumptions and hypotheses without any support. The language aims to hide this, and makes you think about the language instead of its contents. Which is logically unsound: In a sharp peak, micro optimizations would give you a clearer signal where the optimum lies since the gradient is steeper.
- How? I think the same argument applies: If it's changing from loop to loop, seems mutable to me.
- No? It has a lifetime of one loop duration, and is constant during that duration. Seems perfectly fine to me.
- With WSL2, Windows is better, sad, but true.
- Larger breaker and thicker wires!
- > Since the loop copies data pointer by pointer, it can handle the case of overlapping data.
I don't think this loop does the right thing if destination points somewhere into source. It will start overwriting the non-copied parts of source.
- It's niche /for development work/. Being used by a developer doesn't make it used for development. Or the most used developing tool would be the toilet.
- Unfortunately, very true. I think Canada ranks last in the G7 on GDP-percentage science funding. Salaries seem to be OKish, maybe similar to Germany, but cost of living (mainly housing, food is comparable cheap) in Vancouver and Toronto is very high.
UK salaries are atrociously bad. Not sure about funding, but it doesn't seem to be great.
- Honest question on this: It is clear that first amendment protects the rights of protesters from persecution by the government. But does that mean that the government would need to endure protests in any federal building at any time?
If not, I think you could make a case that UCLA could kick protesters out, for example if they take over a building. In contrast to a private university, they probably couldn't act on what the protesters do outside of the university. But I do think that they must have some regulatory power on campus.
- I don't think this is quite correct, at least for private universities. They absolutely have the right to disallow protests on their property. The first amendment protects you from the government -- the government cannot force the university to take these action, but the university could totally do it on their own. It's a little bit murkier with state universities, because you could argue they are a part of the government.
- That's not what people have complained about. People complained about that in the US a government rule seems to be "don't criticize the government".
What about Christian Drosten who criticized a lot of the Corona decisions, Jan Boehmermann who is very critical but still employed by state-financed TV. Fridays for Future?
- Funding is always a lottery. It's not like everyone gets some, but it's less. It's more that less people get funds. So you really have to look at the average.
But you are right. That's why I said: can change quickly.
- My point is this: I do think that some universities gave the protestors way too much leeway. It clearly created a hostile work environment for some - Jewish, and I grant you also Muslim students. The universities responsibility is to keep that in check, and some failed badly. Some did better.
I also want to add my own observation, which might be biased: There was a clear, sizeable fraction of the protests that was beyond "pro-Palestine / anti-Israel's Palestine policy". There was celebration of Hamas and of the attack, especially in the first days.
- Funding levels are lower. Salaries are lower. In physics, from the places I know in the US, about half of the faculty is foreign, and about 30% is German. This is at top universities. That wouldn't be the case if these people would find similar conditions in Europe. But it can change quickly.
- As long as you are not a "Beamte", i.e. a special case of state employee, yes.
- From what I know, some received the death threats because they were Jewish, some because they supported Israel. One of them had the photos of the missing hostages up on the door. They were torn down repeatedly.
- To add some more anec-data: I have a couple of Jewish colleagues. Several of them received death threats. This is not acceptable.
- Also: Even if they could, why would they? Grants are for research. Research only very indirectly affects their income. They could probably accept more students (so more tuition) if they would say to the faculty: no research, more teaching.
An uber driver who gets rich by other means will stop driving for uber, not drive for uber for free.
- Are you sure?
- You are not supposed to reimplement DOOM with templates.
- True, but there are many more people that speak no English, or so badly that an article would be hard to understand. I face this problem now with the classes I teach. It's an electronics lab for physics majors. They have to write reports about the experiments they are doing. For a large fraction, this task is extraordinary hard not because of the physics, but because of writing in English. So for those, LLMs can be a gift from heaven. On the other hand, how do I make sure that the text is not fully LLM generated? If anyone has ideas, I'm all ears.
- Additionally, the total amount of solar and wind is limited. Surely more than we need now, don't get me wrong, but how much more? Factor 2? 10? I could see a future that is extremely energy hungry, and not just because of AI.
- The sun is very close to a black body radiator, so all wavelength. The atmosphere and water filters a lot.
It is actually quite strange that plants are green -- that's the wavelength the atmosphere lets through particularly well, so would be particular good to be absorbed instead of reflected, for energy production. It seems nature hasn't come up with a good, cheap way to move the absorption into that wavelength.
- Not having the data readily available slows it down. Having more random and less well targeted actions hit the supporters, so weakens the support. Is that enough? No. But I still lock my door, even if this will only slow down a determined thief.
Additionally, data collected by the government can also be misused by others. So it's still better to not collect unnecessary.
- True, but there is a large fraction of people who are not like this, but haven't given the dangers of data collection enough thought. You can reach those. Are that enough people? Let's hope so.
I really fear for our older generations and those who are less tech-affine. What chance do they have to not be scammed by AI generated videos, fed by exfiltrated private data of them and their family. Grandparent scam on steroids.
- That what doesn't exist for the license plate dataset? I am sure there are good reasons for having that dataset. For most data collection, there are good reasons.
My argument is that just because we decided that "police with camera" is a worthy trade-off, you cannot use this as an argument for "license plate scanning is a worthy trade-off". It could be that it is, but it doesn't follow from "it's a scaled up version of police with camera".
I would actually argue that only a small percentage of programmers know what happens in code on an instruction level, and near none on a micro-op or register level. Vibe-coding is just one more level of abstraction. The new "code" are the instructions to your LLM.